DESCRIPTION: This is a proposal to address a variety of biostatistical problems motivated by the use of magnetic resonance neuroimaging (MRN) of brain structure and function in the study of psychiatric and neurological outcomes in healthy and diseased subjects, both young and old. Research aims include development and refinement of (1) longitudinal data methods for the analysis of data collected from functional activation and structural volumetric studies; (2) flexible methods of analysis of MRN data based on the newly developing field of functional data analysis; (3) efficient, cost-effective design strategies for studies involving multiple outcomes, including MRN; (4) methods to study MRN outcomes as correlates and predictors of disease severity and onset. Empirical data analysis will play a central role in all aims. Data are available from a variety of sources and encompass examples of conventional structural MRN, functional MRN and magnetic resonance spectroscopic neuroimaging.